[Freegis-list] Paper on the pieces of a free gis.

[Jan-Oliver Wagner]

On Thu, May 11, 2000 at 07:22:07PM -0700, Adrian Vance Custer wrote:
> I've put together some thoughts on how a free GIS could be assembled. It needs
> some fleshing out but it's a starting point. My intent is to build an object
> oriented GIS- I call it gnuGIS- which can be used for wierd kinds of analysis.
> I'm hoping to figure out the object structures, the search tree structure, and
> how to modularize the whole thing through the implementation of API layers.
> Eventually, I'll develop the details of different possible implementations. The
> FMaps project is taking a particular slant towards search structures-putting
> everything in the database- and I'll be interested to see how it goes.

First of all: I have put the document onto our web-server at
ftp://intevation.de/freegis/misc/ so you all can refer to it if you need to.
Adrian, would it be possible for you to put the text as html pages
somewhere on a server? We can place it on the FreeGIS server if you like,
but I would need something else than PDF format.

As you all know GIS is a extremely complex thing. Many facettes have developed
different views on what a GIS, how it should work and what it should deliver.
At the moment I seen several people wishing to create a new, free GIS system
that should be superior to what they use right now (fmaps, OSSIM, GeoPoly
and people like Adrian). I see fresh ideas presented for discussion and
I really appreciate this.
Still there are several GIS packages (GRASS being the most prominent one)
which are very powerful and can already be used to solve problems.
Their drawbacks are often a miserable user interface and old programming
styles (makeing it difficult to work on the source codes).

To my mind there will never be 'the one and only ultimate free GIS'.
A set of applications will exist. In some areas at a competitive level,
in some areas at a cooperative level. Both increases quality of the product.

For the near future I propose to combine the new ideas with the existing
software. Lets discuss what realy needs to be replaced in the existing tools
and what can be incorporated in an intelligent way. My idea is that we
primarily need smart user interface shells abstracting what is used below.
This essentially means that we should make existing stuff act as
(forgive me this fancy word) components.
Adrians object oriented approach could be model for this.


Jan

-- 
Jan-Oliver Wagner				http://intevation.de/~jan/

Intevation GmbH				  	     http://intevation.de/
FreeGIS						       http://freegis.org/